|
Simon Eickhoff
@
INM7_ISN
Neuss, Deutschland
|
|
News & Views on Brain mapping | Machine-Learning | Translational Neuroscience. Director INM-7 @fz_juelich; Professor for Systems Neuroscience @hhu_de.
|
|
|
2.587
Tweetovi
|
739
Pratim
|
4.944
Osobe koje vas prate
|
| Tweetovi |
|
Simon Eickhoff
@INM7_ISN
|
5 min |
|
Locus coeruleus connectivity at 3T based on individual localization using a neuromelanin sensitive sequence as compared to MNI space masks.
Results largely converge, though the TPJ may be driven by other nuclei. Good news for atlas-based analyses!
buff.ly/3b7auYk pic.twitter.com/WswNzQSHA5
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
3 h |
|
Is connectivity-based fMRI neurofeedback nothing but a (very expensive) breathing training?
Given effectiveness of global signal regression motion may also play a role
Unsurprising but convincing demonstration of how susceptible rs-FC is to physiology
sciencedirect.com/science/articl… pic.twitter.com/h1WkY5HuoJ
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
23 h |
|
Informative and entertaining read on how counter-intuitive high dimensional space is, on the improbable effectiveness of #deeplearning and the concept of #ArtificialIntelligence
The easy answer to the unreasonable effectiveness, though? overfitting
pnas.org/content/early/…
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
2. velj |
|
Not really surprising no. But still a nice and relevant illustration. In fact, I had suspected this (80% by simply overfitting) but seeing it demonstrated is still nice
|
||
|
|
||
| Simon Eickhoff proslijedio/la je tweet | ||
|
MHASC ©
@MHASC1
|
31. sij |
|
Happy to advertise a 3yr EU-funded PhD position in Lille (co-supervised w #F_Cabestaing) on #fMRI based #Neurofeedback of #Hallucinations
More on application very soon
#PRT
@echrgroup @IPsychiatrie @HumanBrainProj @INM7_ISN @NeuroSpin_91 @hearingvoice @PhilCorlett1 @isiteULNE
|
||
|
|
||
| Simon Eickhoff proslijedio/la je tweet | ||
|
FZ Jülich-INM7
@fzj_inm7
|
29. sij |
|
Heute hat Edna Cieslik-Köchling den Schülern des Gymnasiums @HOverbach beim Hirnforschungstag im @fz_juelich einen Einblick in die Hirnforschung und unseren Arbeitsbereich der interindividuellen Variabilität des Gehirns gegeben. Eine schöne Möglichkeit, um von Experten zu lernen! pic.twitter.com/MHHeZ7R9HP
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
27. sij |
|
I feel sorry for the second type. Must feel rather lonely without the constant chatter inside your head.
How do they weight arguments? How go through a list? How do they rehearse discussions?
All "narrative" person, best conversations happen inside!
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
26. sij |
|
Absolutely agreed on the motivations, very good advice. Will try to keep that in mind, but..
.. There is still a long way to go. If you haven't made a public announcement to stop flying but rather quietly reduce, answers from organizers are often unenthusiastic. Needs to change
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
26. sij |
|
Native English advantage becomes particularly relevant at "higher" journals with a strict word limit (and expectation of ample results)
Few would acknowledge this, though, and attribute successful submissions to their great science and clarity of presentation
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
24. sij |
|
It's part of a series of slides. Here the focus is on univariate vs multivariate effects. Next slide: out-of-sample validation to prevent overfitting
|
||
|
|
||
| Simon Eickhoff proslijedio/la je tweet | ||
|
Thomas Yeo
@bttyeo
|
24. sij |
|
|
||
|
|
||
| Simon Eickhoff proslijedio/la je tweet | ||
|
Cognition and Brain Sciences Unit
@mrccbu
|
23. sij |
|
Delighted to have Prof. Simon Eickhoff (Institute of Systems Neuroscience, U. of Dusseldorf) visit the @mrccbu, & give a talk at today’s #Chaucerclub on:
“Combining brain mapping with machine-learning on individual differences”
#brainmapping #machinelearning @INM7_ISN @HHU_de pic.twitter.com/IGlO2hH0Hx
|
||
|
|
||
| Simon Eickhoff proslijedio/la je tweet | ||
|
Dr Camilla Nord
@camillalnord
|
23. sij |
|
Sumo wrestlers and basketball players ⛹️♀️ may be statistically indistinguishable (p=0.1) if you compare their weights or height, but multivariable approach (height&weight) has >90% classification accuracy. This is an essential lesson for psychiatric neuroimaging! @INM7_ISN @mrccbu pic.twitter.com/iCUShT7NFK
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
22. sij |
|
Indeed, functional characterization is per se a complex endeavor as we habe written here: pubmed.ncbi.nlm.nih.gov/29501326-how-t…
Hence we never claim and reverse inference or exclusivity
Yet, showing that a region robustly activates by pain stimuli is a critical to understanding pain circuits
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
22. sij |
|
“Using a term like nonlinear function is like referring to the bulk of zoology as the study of non-elephant animals.”
– Stanislaw Ulam pic.twitter.com/xkMN2gJDyZ
|
||
|
|
||
| Simon Eickhoff proslijedio/la je tweet | ||
|
Journals of The American Psychiatric Association
@APAPubJournals
|
22. sij |
|
The finding of common neurocircuit substrates for the cognitive and emotional impairment prevalent across disorders could advance multimodal therapeutics and reconceptualize mental illness. bit.ly/2TtYWbb @MUSC_COM @INM7_ISN pic.twitter.com/d6Z37jM1Uh
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
22. sij |
|
How about posting preprints of those works that are currently in review? This would provide the benefit of showing that the work is finished while incurring little risk
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
22. sij |
|
Happy to curate the account for a week, talking about neuroimaging and machine-learning, the individualized prediction of mental traits and potential clinical applications...
Plus research in a multidisciplinary @helmholtz_de center
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
22. sij |
|
Sehr gerne!
|
||
|
|
||
|
Simon Eickhoff
@INM7_ISN
|
22. sij |
|
Seeing "in preparation" frequently and wondering what others think about these was actually a key motivation for this poll.
On the flip-side, many appointed committees (in Germany) still only consider published papers...
|
||
|
|
||